// ========================================================================================
static void BuildBsonDocument(List <string> parsedText, string crawledURL)
{
List <BsonDocument> dataDocuments = new List <BsonDocument>();
// Dictionary containing keywords desired, and a list of all contexts in which they were used
Dictionary <string, Pair <int, List <string> > > contextCache = DataScraper.GetWordCountAndContext(parsedText, Constants.DefaultIgnoreWordsTXT);
//make list from dict to sort
var dictList = contextCache.ToList();
// Sort takes a comparison operator
// Comparison(x,y) -> less than 0 if x < y, 0 if equal, greater than 0 if x > y
// for all keyValuePairs in dict, sort based on the frequency count
// pair: word : list of
dictList.Sort((pair1, pair2) => pair1.Value.Item1 > pair2.Value.Item1 ? -1 : 1);
var sentimentAnalysis = new BsonDocument();
var numWords = dictList.Count > 100 ? 100 : dictList.Count;
for (int i = 0; i < numWords; i++)
{
//Log.Logger.Debug("Getting Context words for " + dictList[i].Key);
if (dictList[i].Key == "")
{
continue; // Skips the stupid empty string keyword problem we havn't fixed yet...
}
// Excludes words we don't care about
var desiredWords = DataScraper.ExcludeWords(dictList[i].Value.Item2);
//the context sentences
//number of occurances of context words for a given keyword
var contextWordCount = DataScraper.GetWordCount(desiredWords);
sentimentAnalysis.Add(new BsonElement(
dictList[i].Key, new BsonDocument
{
{ "Count", dictList[i].Value.Item1 },
{ "ContextSentences", new BsonArray(dictList[i].Value.Item2) },
{ "ContextWordFrequency", new BsonDocument(contextWordCount) }
}
));
}
// BSON doc
var bson = new BsonDocument
{
{ "WebsiteTitle", DataScraper.siteTitle },
{ "URL", crawledURL },
// {"Raw", rawPageText},
{ "SentimentAnalysis", sentimentAnalysis }
};
if (bson != null)
{
DataScraper.dataDocuments.Add(bson);
}
}
// PRIVATE CLASS MEMBERS
// MAIN ===================================================================================
/// <param name="args"> Command line arguements passed to executable. </param>
static async Task Main(string[] args)
{
// Creates a Logger object from Serilog. Writes up to Debug level prints.
SetupLogger();
Log.Logger.Information("Darkweb Data Scraper start...");
// Parses command line arguements and stores them in "parsedArgs"
SetupParser(args);
// I made this function to move this setup out of main.
// Returns a TorSharpSettings object for use with TorSharp.
var settings = SetupTorSharpSettings();
// Idk exactly how this works but like... its for torsharp
// its uh... setting up torsharp "tools"...also its asyncronous
await SetupTorSharpTools(settings);
// * starts tor proxy -----------------------------------------------------------------
using (var proxy = new TorSharpProxy(settings))
{
var waiting = true;
while (waiting)
{
// block untill we wait for TorSharp Proxy to be configured
await proxy.ConfigureAndStartAsync();
waiting = false;
}
// * SETUP AND EXECUTE CRAWLER ================================================
// Setup Crawler configuration
CrawlConfigurationX crawlConfig = new CrawlConfigurationX
{
MaxPagesToCrawl = 30, // Max total urls this crawler should crawl
MaxCrawlDepth = 1, // Depth for crawler to traverse urls
IsJavascriptRenderingEnabled = false, // Should crawler render JS?
JavascriptRenderingWaitTimeInMilliseconds = 2000, // How long to wait for js to process
MaxConcurrentSiteCrawls = 1, // Only crawl a single site at a time
MaxRetryCount = 3 // Retries to connect and crawl site 'x' times
};
if (parsedArgs.InputFile == null) // THIS IS "-s" or "--single"
{
var handler = new HttpClientHandler
{
Proxy = new WebProxy(new Uri("http://localhost:" + settings.PrivoxySettings.Port))
};
// Crawl
await DataScraper.Crawl(crawlConfig, handler, parsedArgs.handlerType, parsedArgs.StartingUri);
BuildBsonDocument(DataScraper.allParsedText, parsedArgs.StartingUri);
//reset vals for next crawl
DataScraper.allParsedText = new List <string>();
DataScraper.siteTitle = "";
DataScraper.dataDocuments = new List <BsonDocument>();
if (MONGO_URI == "")
{
Log.Logger.Information("Database information is no longer accessible or available." +
"You will need to provide your own Mongo details in \"Crawler.cs\".");
}
}
else // THIS IS "-m" or "--multi"
{
string inputFilePath = @parsedArgs.InputFile;
var sitesToCrawl = GenerateSiteList(inputFilePath);
for (int i = 0; i < sitesToCrawl.Count; i++)
{
var handler = new HttpClientHandler
{
Proxy = new WebProxy(new Uri("http://localhost:" + settings.PrivoxySettings.Port))
};
// Crawl
await DataScraper.Crawl(crawlConfig, handler, parsedArgs.handlerType, sitesToCrawl[i]);
BuildBsonDocument(DataScraper.allParsedText, sitesToCrawl[i]);
if (MONGO_URI == "")
{
Log.Logger.Information("Database information is no longer accessible or available." +
"You will need to provide your own Mongo details in \"Crawler.cs\".");
}
if (MONGO_URI != "" && MONGO_COLLECTION_NAME != "" && MONGO_DATABASE_NAME != "")
{
var client = new MongoClient(MONGO_URI);
var database = client.GetDatabase(MONGO_DATABASE_NAME);
var collection = database.GetCollection <BsonDocument>(MONGO_COLLECTION_NAME);
collection.InsertMany(DataScraper.dataDocuments);
}
//reset vals for next crawl
DataScraper.allParsedText = new List <string>();
DataScraper.siteTitle = "";
DataScraper.dataDocuments = new List <BsonDocument>();
}
}
// * ==========================================================================
// Stop the TorSharp tools so that the proxy is no longer listening on the configured port.
proxy.Stop();
}
// * ----------------------------------------------------------------------------------
}
